Privacy protection data exchange method and system under zero-trust network architecture

Abstract

The invention belongs to the field of privacy protection data, and provides a privacy protection data exchange method and system under a zero-trust network architecture in order to solve the problem that a traditional scheme cannot cope with the privacy leakage risk through data analysis. The privacy protection data exchange method under the zero-trust network architecture comprises the following steps: performing tagging processing on user behavior data, equipment state data and operation environment data; executing multi-dimensional dynamic risk assessment, and calculating a comprehensive risk score; adjusting privacy budget and noise intensity according to a comprehensive risk scoring result, and protecting a data query result in combination with an adaptive differential privacy mechanism; performing dynamic data desensitization based on information entropy on a data query result; embedding a self-evolved data watermark into the data after the data query result desensitization processing is completed; the embedded watermark data is exchanged through the encrypted transmission channel, and the transmission process is monitored in real time, so that the security and integrity of the data exchange under the zero-trust network architecture are ensured.

Description

Technical Field

[0001] The present invention belongs to the field of privacy protection data, and in particular, relates to a privacy protection data exchange method and system under a zero-trust network architecture. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] In an open network environment, data exchange faces serious risks of privacy leakage and data tampering. Traditional security measures (such as encryption and access control) cannot effectively prevent attackers from inferring sensitive information from data or tampering with data during transmission. Especially under the zero-trust architecture, all data exchanges must undergo strict identity authentication and continuous monitoring, but existing solutions still have deficiencies in privacy protection and anti-tampering, mainly manifested in the following aspects: Traditional solutions cannot cope with the risk of privacy leakage through data analysis; data tampering during transmission is difficult to effectively detect and prevent; existing solutions have weak defense capabilities when dealing with complex attacks (such as tampering, reconstruction, etc.). Summary of the invention

[0004] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a privacy-preserving data exchange method and system under a zero-trust network architecture, which ensures privacy protection and integrity in the data exchange process through multi-dimensional dynamic risk assessment, adaptive differential privacy, dynamic desensitization and watermark embedding technology, and meets the high security requirements under the zero-trust architecture.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: A first aspect of the present invention provides a privacy-preserving data exchange method under a zero-trust network architecture.

[0006] A privacy protection data exchange method under a zero-trust network architecture, comprising: Obtain user behavior data, device status data, and operating environment data and label them; Perform multi-dimensional dynamic risk assessment on the labeled data and calculate the comprehensive risk score; The privacy budget and noise intensity are adjusted according to the comprehensive risk score results, and the adaptive differential privacy mechanism is combined to protect the data query results. Perform dynamic data desensitization based on information entropy on the data query results; Embed a self-evolving data watermark in the data after the data query result is desensitized; The embedded watermark data is exchanged through an encrypted transmission channel, and the transmission process is monitored in real time to ensure the security and integrity of data exchange under the zero-trust network architecture.

[0007] As an implementation method, the calculation formula of the comprehensive risk score is: ; in, Indicates the device The composite risk score at time t; , , , is the weight coefficient; Indicates the device behavioral risk score at time t; represents the environmental risk score at time t; Indicates the device Historical credit risk score at time t; Indicates the device The abnormal risk score at time t.

[0008] As an embodiment, the device Behavioral risk score at time t The expression is:

[0009] in: is the weight of the j-th operation behavior; For equipment The score of the jth action performed at time t; n represents the total number of action types; is the time decay factor, ensuring that the latest operational behavior has a greater impact on the evaluation; The time when the operation occurs.

[0010] As an implementation method, the environmental risk score at time t The expression is:

[0011] in: and is the weight of geographic location and network status, and ; is the risk score of the geographical location; is the network status score.

[0012] As an embodiment, the device Historical credit risk scores at time t The expression is:

[0013] in: Indicates the device The number of safe states; Indicates the device The number of abnormal states; equipment Abnormal risk score at time t The expression is:

[0014] in: and are the weight coefficients of the number of abnormal behaviors and the device status respectively; Score the current abnormal status of the device.

[0015] As an implementation method, the expression of the adaptive differential privacy mechanism to protect the data query result is:

[0016]

[0017] in: is the query result after adding noise; is the original query function; is the sensitivity of the query function; Budget for privacy; ( ) represents Laplace transform; budget for initial privacy; is the risk factor weight; is the comprehensive risk score of the device.

[0018] As an implementation method, in the process of performing dynamic data desensitization based on information entropy on the data query results: When the information entropy H(d) of data d is less than the preset first threshold When , there is no desensitization processing and the original data d is directly output; When the information entropy H(d) of data d is greater than or equal to the preset second threshold , the data is completely masked, i.e. all sensitive information is obscured; Greater than ; When the information entropy H(d) of data d is between the threshold closed interval [ , ], partial desensitization is implemented, and the degree of desensitization is , the value is between 0 and 1, the closer to 1, the more masking; in, ; H(d) is the information entropy of data d; p(x) is the probability of occurrence of data item x; For data items .

[0019] As an implementation method, the expression of the data watermark embedded in the self-evolving data is:

[0020] in: Represents data watermark data embedded in self-evolving data; Represents the original data; Represents the watermark information embedded in the original data; is the watermark optimized by genetic algorithm, indicating the optimal watermark design selected in the attack mode; is the watermark strength coefficient, which controls the impact of the watermark on the original data, and 0≤ ≤1.

[0021] As an implementation method, the optimal watermark design is evaluated using a fitness function, wherein the fitness function formula is:

[0022] in: represents the fitness function; The watermark is in the current attack mode The robustness under ,is quantified by the normalized cross-correlation method; To make the watermark imperceptible, peak signal-to-noise ratio quantization is used; is the constant weight coefficient.

[0023] A second aspect of the present invention provides a privacy-preserving data exchange system under a zero-trust network architecture.

[0024] A privacy protection data exchange system under a zero-trust network architecture, comprising: A label processing module is used to obtain user behavior data, device status data and operating environment data and label them; Dynamic risk assessment module, which is used to perform multi-dimensional dynamic risk assessment on the labeled data and calculate the comprehensive risk score; The differential privacy protection module is used to adjust the privacy budget and noise intensity according to the comprehensive risk score results, and combines the adaptive differential privacy mechanism to protect the data query results; A dynamic data desensitization module is used to perform dynamic data desensitization based on information entropy on the data query results; A self-evolving watermark embedding module, which is used to embed a self-evolving data watermark into the data after the data query result desensitization processing is completed; The data encryption transmission module is used to exchange embedded watermark data through an encrypted transmission channel and monitor the transmission process in real time to ensure the security and integrity of data exchange under the zero-trust network architecture.

[0025] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention performs multi-dimensional dynamic risk assessment on labeled user behavior data, device status data, and operating environment data, and can dynamically adjust security policies based on factors such as device behavior, credit, and environment, thereby improving security during data exchange.

[0026] (2) The present invention adjusts the privacy budget and noise intensity according to the comprehensive risk scoring results. It also effectively prevents sensitive data from being speculated or leaked during the exchange process by integrating adaptive differential privacy mechanism and dynamic data desensitization technology, thereby ensuring data privacy.

[0027] (3) The present invention uses self-evolving data watermark technology to embed a self-evolving data watermark into the data after the data query result is desensitized, and monitors the integrity of the data in real time to prevent the data from being tampered with or forged during transmission.

[0028] (4) The present invention meets the high security requirements of the zero-trust network architecture, realizes the secure exchange of data in a complex network environment, is suitable for a variety of scenarios with high privacy protection requirements, can provide full-process monitoring and auditing functions, detect and prevent abnormal operations in real time, and ensure that the data exchange process is transparent, traceable and compliant.

[0029] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0031] Figure 1 It is a flow chart of a privacy protection data exchange method under a zero-trust network architecture according to an embodiment of the present invention; Figure 2 It is a structural diagram of a privacy protection data exchange system under a zero-trust network architecture according to an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0033] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0034] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0035] Terminology explanation: Zero Trust Architecture is a network security model that emphasizes continuous monitoring and evaluation based on the principle of "never trust, always verify". In the Zero Trust model, no matter where the request comes from, only verified devices can access the corresponding resources, thereby reducing security risks and improving overall network security.

[0036] Figure 1 This is a flow chart of a privacy protection data exchange method under a zero-trust network architecture according to an embodiment of the present invention. Figure 1 , an embodiment of the present invention provides a privacy protection data exchange method under a zero-trust network architecture, comprising: S101: Obtain user behavior data, device status data, and operating environment data and label them.

[0037] During the specific implementation process, the system log function is turned on to record all data processing and exchange operations, including parameter settings, data transmission and security events in each step; the audit system is enabled to track key processes and operations in data exchange to ensure that system operations can be audited for subsequent analysis and compliance checks.

[0038] The system needs to comprehensively collect risk perception data related to data exchange and label these data according to preset classification standards; through labeling, the system can implement corresponding security controls and privacy protection measures on different types of data in subsequent steps, thereby improving the security and reliability of data exchange; and update data labels in real time.

[0039] Automatically collect the specific operations performed by users or devices during data exchange, namely user behavior data, which includes operations such as reading, writing, modifying, and deleting data. Each piece of operation data includes the type of operation, the time of occurrence, the frequency of the operation, and the specific data objects involved.

[0040] The samples of user behavior data include: operation type label, operation frequency label and operation time period label.

[0041] Operation type labels: such as "read", "write", "modify", and "delete". These labels can help the system distinguish different types of operations and assign different weights to different types of operations in subsequent risk assessments.

[0042] Operation frequency label: Labels the operation behavior as "high frequency", "medium frequency" or "low frequency" according to the operation frequency. For example, frequent read or write operations in a short period of time may be abnormal behavior, and the system will perform additional risk assessment on such operations through frequency labels.

[0043] Operation time period label: Label the operation as "working hours" or "non-working hours" based on the time the operation occurred. Operations performed during non-working hours may have higher risks and therefore will be assigned a higher risk weight.

[0044] Among them, the device status data includes the device security status and network connection status.

[0045] The labels of the device status data include the device security status label and the network connection status label; Device security status label: such as "secure" or "abnormal". When a device is marked as "abnormal", there may be security vulnerabilities (such as not installing the necessary patches), and these devices will receive higher security protection in subsequent steps.

[0046] Network connection status label: such as "Secure Network" or "Public Network". When a device is connected through a public network, the risk will increase significantly, and the system will strengthen data protection through encryption, privacy protection, etc.

[0047] The operating environment data includes external environment information such as the geographic location during operation and the stability of network connection. The system collects the current geographic location of the device through GPS, IP address, etc., and monitors network performance indicators such as network delay and packet loss rate to ensure the security and stability of the operating environment.

[0048] The labels of the operating environment data include geographic location labels and network status labels; Geographic location labels: such as "trusted location" or "unauthorized location". When a device is in an unauthorized geographic location (such as in a public place), its operation will be marked as a high-risk operation and trigger additional security checks.

[0049] Network status label: such as "stable network" or "unstable network". When the network connection status is unstable, the security of data transmission will be threatened. The system will encrypt such network environments or add privacy protection measures.

[0050] S102: Perform multi-dimensional dynamic risk assessment on the labeled data and calculate a comprehensive risk score.

[0051] In the specific implementation process, the calculation formula for the comprehensive risk score is: ; in, Indicates the device The composite risk score at time t; , , , is the weight coefficient; Indicates the device behavioral risk score at time t; represents the environmental risk score at time t; Indicates the device Historical credit risk score at time t; Indicates the device The abnormal risk score at time t.

[0052] The weight parameter here , which is used to adjust the influence of each dimension (such as behavioral risk, historical credit risk, environmental risk and abnormal risk) in the multi-dimensional dynamic risk assessment model. The setting of each weight should be adjusted according to the device type and security requirements involved in the data exchange.

[0053] Among them, equipment Behavioral risk score at time t The expression is:

[0054] in: is the weight of the j-th operation behavior; For equipment The score of the jth action performed at time t; n represents the total number of action types; is the time decay factor, ensuring that the latest operational behavior has a greater impact on the evaluation; The time when the operation occurs.

[0055] For different operation type labels, weights Make different adjustments: For the "read" operation, the weight is set to the baseline value; for the "write" operation, the weight is increased by 20%; for the "modify" operation, the weight is increased by 10%; for the "modify" operation, the weight is increased by 10%.

[0056] The weights are adjusted differently for different operation frequency type labels: For "high frequency" operations, the weight is increased by 50%; for "medium frequency" operations, the weight is increased by 25%; for "low frequency" operations, the weight is reduced by 20%.

[0057] According to different operation time period labels, behavioral risks Make different adjustments: During "non-working hours", behavioral risks increase by 30%; during working hours, they are at the baseline value.

[0058] Environmental risk score at time t The expression is:

[0059] in: and is the weight of geographic location and network status, and ; is the risk score of the geographical location; For example, when a device is in a "trusted location", the risk of the geographic location is low. ; When the device is in an "unauthorized location", the risk of the geographic location is higher, ; When the device is connected to a "safe network", the network status is good, , when the device is connected to the "public network", the network status is poor, .

[0060] equipment Historical credit risk scores at time t The expression is:

[0061] in: Indicates the device The number of safe states; Indicates the device The number of abnormal states; equipment Abnormal risk score at time t The expression is:

[0062] in: and are the weight coefficients of the number of abnormal behaviors and the device status respectively; Score the current abnormal state of the device. For example, when the device status is "safe", the abnormal state score is low. ,When the device status is "abnormal", the abnormal status score is high, .

[0063] S103: Adjust the privacy budget and noise intensity based on the comprehensive risk scoring results, and use the adaptive differential privacy mechanism to protect data query results.

[0064] In the specific implementation process, the expression of the adaptive differential privacy mechanism to protect the data query results is:

[0065]

[0066] in: is the query result after adding noise; is the original query function; is the sensitivity of the query function, which represents the most likely single change value in the query; is the privacy budget. The larger the privacy budget, the less noise is added. ( ) represents Laplace transform; budget for initial privacy; is the risk factor weight, which is used to control the impact of risk on the privacy budget; is the comprehensive risk score of the device. The higher the risk, the lower the privacy budget.

[0067] In this embodiment, the adaptive differential privacy mechanism combines the results of dynamic risk assessment and adjusts the privacy budget. The value controls the intensity of the noise.

[0068] Among them, when the comprehensive risk of the equipment When it is higher than 0.8, it is considered a high-risk device and the privacy budget ; When the comprehensive risk of a device is between 0.5 and 0.8, it is considered a medium-risk device and the privacy budget ; When the overall risk of the device is less than 0.5, the privacy budget .

[0069] Initial privacy budget here Used to control the privacy protection strength when querying data, the default value is 0.5. Values ​​of will result in higher privacy protection (i.e. more noise) but may reduce the usefulness of the data. This parameter is usually set based on the sensitivity of the data and the complexity of the expected queries.

[0070] S104: Perform dynamic data desensitization based on information entropy on the data query results. Information entropy is used to measure the sensitivity of data, and data desensitization is adjusted according to the level of information entropy to ensure that data under different security requirements are reasonably protected.

[0071] Specifically, in the process of performing dynamic data desensitization based on information entropy on the data query results: When the information entropy H(d) of data d is less than the preset first threshold When , there is no desensitization processing and the original data d is directly output; When the information entropy H(d) of data d is greater than or equal to the preset second threshold , the data is completely masked, i.e. all sensitive information is obscured; Greater than ; When the information entropy H(d) of data d is between the threshold closed interval [ , ], partial desensitization is implemented, and the degree of desensitization is , the value is between 0 and 1, the closer to 1, the more masking; in, ; H(d) is the information entropy of data d; p(x) is the probability of occurrence of data item x; For data items .

[0072] Here is the dynamic desensitization threshold based on information entropy , the information entropy value is lower than Data of can be left un-anonymous or lightly anonymized; and Partial desensitization is performed if the The data is completely desensitized; the threshold and It should be adjusted dynamically based on the actual use of the data, privacy needs and expert opinions.

[0073] Data with higher information entropy will be more strictly desensitized, while data with lower information entropy will retain more original information, thus achieving a balance between privacy protection and data availability.

[0074] S105: Embed a self-evolving data watermark into the data after the data query result desensitization processing is completed.

[0075] The expression of the data watermark embedded in the self-evolving data is:

[0076] in: Represents data watermark data embedded in self-evolving data; Represents the original data; Represents the watermark information embedded in the original data; is the watermark optimized by genetic algorithm, indicating the optimal watermark design selected in the attack mode; is the watermark strength coefficient, which controls the impact of the watermark on the original data, and 0≤ ≤1.

[0077] The optimal watermark design is evaluated using a fitness function, where the formula for the fitness function is:

[0078] in: represents the fitness function; The watermark is in the current attack mode The robustness under the condition of ,is quantified by the normalized cross-correlation method. The higher the robustness, the stronger the watermark's resistance to tampering; To ensure the imperceptibility of the watermark, the peak signal-to-noise ratio is used for quantification. The higher the value, the closer the watermark is to the original data and the harder it is to detect. is the constant weight coefficient.

[0079] In this embodiment, the self-evolving watermark technology is combined with a genetic algorithm to enable the watermark design to be adaptively optimized when dealing with different attack modes, thereby enhancing the watermark's anti-tampering capability.

[0080] Initially set the key parameters of the self-evolving data watermarking technology, including the watermark embedding method, strength control, and the initial design of the watermark; configure the genetic algorithm parameters used for watermark evolution, such as population size, crossover probability, mutation probability, evolution algebra, etc.; these parameters determine the speed and effect of watermark evolution, ensuring the concealment and robustness of the watermark.

[0081] The watermark design can be continuously optimized according to different attack modes (such as tampering, reconstruction, replication, etc.). The genetic iteration process is as follows: Initial population generation: Generate a set of initial watermark designs (population).

[0082] Fitness evaluation: Each watermark design is evaluated based on its robustness and imperceptibility.

[0083] Selection and crossover: Select the optimal watermark design according to the fitness function, perform crossover combination, and generate a new watermark.

[0084] Mutation: Introducing random changes to diversify the watermark design and increase its ability to resist attacks.

[0085] Evolutionary iteration: Repeat the above steps until the optimal watermark design is found .

[0086] It should be noted here that, in other embodiments, the self-evolving watermark technology can also be combined with other optimization algorithms (such as ant colony algorithm, etc.) to enable the watermark design to be adaptively optimized when dealing with different attack modes, thereby enhancing the watermark's anti-tampering capability.

[0087] After the optimal watermark design is dynamically generated, it is embedded into the data to ensure the tamper-proof and concealment of the data during transmission and storage.

[0088] S106: Exchange the embedded watermark data through an encrypted transmission channel and monitor the transmission process in real time to ensure the security and integrity of data exchange under the zero-trust network architecture.

[0089] Select and configure the encryption standards used in transmission (including but not limited to AES, RSA) to ensure the encryption of all transmitted data; set access control policies and permission rules to ensure that only authenticated and authorized users can access and configure system resources.

[0090] Before data is exchanged, the system authenticates and authorizes the recipient to ensure that only legitimate users or devices can receive data.

[0091] Data is transmitted through encrypted channels (such as TLS / SSL) to prevent data from being intercepted or tampered with during transmission and ensure data confidentiality.

[0092] The processed data, including the desensitized, privacy-protected and watermark-embedded parts, begins to be transmitted to the recipient through an encrypted channel to ensure transmission security.

[0093] After receiving the data, the receiver verifies the integrity of the data through the watermark to ensure that the data has not been tampered with or damaged.

[0094] The entire exchange process is monitored in real time by the system, and every step of the operation is recorded. Any abnormal behavior or unauthorized access will trigger a warning and record, ensuring that data transmission is traceable.

[0095] Once abnormal behavior is detected, the system will immediately interrupt transmission, issue an alarm and record the event to prevent data leakage or tampering.

[0096] After the data transmission is completed, the system performs a final watermark and privacy protection check to confirm the integrity and security of the data.

[0097] Figure 2 A schematic diagram of the structure of a privacy protection data exchange system under a zero-trust network architecture according to an embodiment of the present invention is given. Figure 2 As shown, an embodiment of the present invention provides a privacy protection data exchange system under a zero-trust network architecture, including: The labeling processing module 201 is used to obtain user behavior data, device status data and operating environment data and perform labeling processing on them; A dynamic risk assessment module 202 is used to perform multi-dimensional dynamic risk assessment on the labeled data and calculate a comprehensive risk score; A differential privacy protection module 203, which is used to adjust the privacy budget and noise intensity according to the comprehensive risk scoring results, and to protect the data query results in combination with an adaptive differential privacy mechanism; A dynamic data desensitization module 204, which is used to perform dynamic data desensitization based on information entropy on the data query results; A self-evolving watermark embedding module 205, which is used to embed a self-evolving data watermark into the data after the data query result desensitization processing is completed; The data encryption transmission module 206 is used to exchange embedded watermark data through an encrypted transmission channel and monitor the transmission process in real time to ensure the security and integrity of data exchange under the zero-trust network architecture.

[0098] It should be noted here that the various modules in the privacy protection data exchange system under the zero-trust network architecture of the embodiment of the present invention correspond one-to-one to the various steps in the privacy protection data exchange method under the above-mentioned zero-trust network architecture, and the specific implementation process is the same, which will not be described in detail here.

[0099] This embodiment constructs an effective privacy-preserving data exchange method, which fully guarantees the privacy, integrity and security of data during the exchange process. This method not only meets the high security requirements under the zero-trust network architecture, but also provides real-time monitoring and auditing functions, effectively preventing abnormal operations and potential risks, and ensuring the compliance and reliability of data exchange. Through the implementation of the present invention, the demand for secure data exchange in an open network environment is fully met.

[0100] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication part, and / or installed from a removable medium. When the computer program is executed by a central processing unit, various functions defined in the device of the present application are executed.

[0101] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products of the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of the processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0102] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A privacy protection data exchange method under a zero-trust network architecture, characterized in that: include: Obtain user behavior data, device status data, and operating environment data and label them; Perform multi-dimensional dynamic risk assessment on the labeled data and calculate the comprehensive risk score; The privacy budget and noise intensity are adjusted according to the comprehensive risk score results, and the adaptive differential privacy mechanism is combined to protect the data query results. Perform dynamic data desensitization based on information entropy on the data query results; Embed a self-evolving data watermark in the data after the data query result is desensitized; The embedded watermark data is exchanged through an encrypted transmission channel, and the transmission process is monitored in real time to ensure the security and integrity of data exchange under the zero-trust network architecture.

2. The privacy protection data exchange method under the zero-trust network architecture according to claim 1, characterized in that: The formula for calculating the comprehensive risk score is: ； in, Indicates the device The composite risk score at time t; , , , is the weight coefficient; Indicates the device behavioral risk score at time t; represents the environmental risk score at time t; Indicates the device Historical credit risk score at time t; Indicates the device The abnormal risk score at time t.

3. The privacy protection data exchange method under the zero-trust network architecture as claimed in claim 2, characterized in that: equipment Behavioral risk score at time t The expression is: in: is the weight of the j-th operation behavior; For equipment The score of the jth action performed at time t; n represents the total number of action types; is the time decay factor, ensuring that the latest operational behavior has a greater impact on the evaluation; The time when the operation occurs.

4. The privacy protection data exchange method under the zero-trust network architecture as claimed in claim 2, characterized in that: Environmental risk score at time t The expression is: in: and is the weight of geographic location and network status, and ; is the risk score of the geographical location; is the network status score.

5. The privacy protection data exchange method under the zero-trust network architecture as claimed in claim 2, characterized in that: equipment Historical credit risk scores at time t The expression is: in: Indicates the device The number of safe states; Indicates the device The number of abnormal states; equipment Abnormal risk score at time t The expression is: in: and are the weight coefficients of the number of abnormal behaviors and the device status respectively; Score the current abnormal status of the device.

6. The privacy protection data exchange method under the zero-trust network architecture as claimed in claim 1, characterized in that: The expression of the adaptive differential privacy mechanism to protect data query results is: in: is the query result after adding noise; is the original query function; is the sensitivity of the query function; Budget for privacy; ( ) represents Laplace transform; budget for initial privacy; is the risk factor weight; is the comprehensive risk score of the device.

7. The privacy protection data exchange method under the zero-trust network architecture as claimed in claim 1, characterized in that: In the process of performing dynamic data desensitization based on information entropy on the data query results: When the information entropy H(d) of data d is less than the preset first threshold When , there is no desensitization processing and the original data d is directly output; When the information entropy H(d) of data d is greater than or equal to the preset second threshold , the data is completely masked, i.e. all sensitive information is obscured; Greater than ; When the information entropy H(d) of data d is between the threshold closed interval [ , ], partial desensitization is implemented, and the degree of desensitization is , the value is between 0 and 1, the closer to 1, the more masking; in, ; H(d) is the information entropy of data d; p(x) is the probability of occurrence of data item x; For data items .

8. The privacy protection data exchange method under the zero-trust network architecture as claimed in claim 1, characterized in that: The expression of the data watermark embedded in the self-evolving data is: in: Represents data watermark data embedded in self-evolving data; Represents the original data; Represents the watermark information embedded in the original data; is the watermark optimized by genetic algorithm, indicating the optimal watermark design selected in the attack mode; is the watermark strength coefficient, which controls the impact of the watermark on the original data, and 0≤ ≤1.

9. The privacy protection data exchange method under the zero-trust network architecture as claimed in claim 8, characterized in that: The optimal watermark design is evaluated using a fitness function, where the formula for the fitness function is: in: represents the fitness function; The watermark is in the current attack mode The robustness under ,is quantified by the normalized cross-correlation method; To make the watermark imperceptible, peak signal-to-noise ratio quantization is used; is the constant weight coefficient.

10. A privacy protection data exchange system under a zero-trust network architecture, characterized in that: include: A label processing module is used to obtain user behavior data, device status data and operating environment data and label them; Dynamic risk assessment module, which is used to perform multi-dimensional dynamic risk assessment on the labeled data and calculate the comprehensive risk score; The differential privacy protection module is used to adjust the privacy budget and noise intensity according to the comprehensive risk score results, and combines the adaptive differential privacy mechanism to protect the data query results; A dynamic data desensitization module is used to perform dynamic data desensitization based on information entropy on the data query results; A self-evolving watermark embedding module, which is used to embed a self-evolving data watermark into the data after the data query result desensitization processing is completed; The data encryption transmission module is used to exchange embedded watermark data through an encrypted transmission channel and monitor the transmission process in real time to ensure the security and integrity of data exchange under the zero-trust network architecture.

Images